Save file id for all fsspec connectors if present

Unstructured-IO · Jul 18, 2024 · ee254c3 · ee254c3
1 parent 5d38703
commit ee254c3
Showing 1 changed file with 3 additions and 0 deletions.
diff --git a/unstructured/ingest/v2/processes/connectors/fsspec/fsspec.py b/unstructured/ingest/v2/processes/connectors/fsspec/fsspec.py
@@ -187,6 +187,9 @@ def get_metadata(self, path: str) -> DataSourceMetadata:
             "protocol": self.index_config.protocol,
             "remote_file_path": self.index_config.remote_url,
         }
+        file_stat = self.fs.stat(path=path)
+        if file_id := file_stat.get("id"):
+            record_locator["file_id"] = file_id
         if metadata:
             record_locator["metadata"] = metadata
         return DataSourceMetadata(